xref: /redis-3.2.3/deps/jemalloc/doc/jemalloc.html (revision 7383c3b1)
1<html><head><meta http-equiv="Content-Type" content="text/html; charset=ISO-8859-1"><title>JEMALLOC</title><meta name="generator" content="DocBook XSL Stylesheets V1.76.1"></head><body bgcolor="white" text="black" link="#0000FF" vlink="#840084" alink="#0000FF"><div class="refentry" title="JEMALLOC"><a name="id286949159"></a><div class="titlepage"></div><div class="refnamediv"><h2>Name</h2><p>jemalloc &#8212; general purpose memory allocation functions</p></div><div class="refsect1" title="LIBRARY"><a name="library"></a><h2>LIBRARY</h2><p>This manual describes jemalloc 3.2.0-0-g87499f6748ebe4817571e817e9f680ccb5bf54a9.  More information
2    can be found at the <a class="ulink" href="http://www.canonware.com/jemalloc/" target="_top">jemalloc website</a>.</p></div><div class="refsynopsisdiv" title="SYNOPSIS"><h2>SYNOPSIS</h2><div class="funcsynopsis"><pre class="funcsynopsisinfo">#include &lt;<code class="filename">stdlib.h</code>&gt;
3#include &lt;<code class="filename">jemalloc/jemalloc.h</code>&gt;</pre><div class="refsect2" title="Standard API"><a name="id286901505"></a><h3>Standard API</h3><table border="0" summary="Function synopsis" cellspacing="0" cellpadding="0" class="funcprototype-table"><tr><td><code class="funcdef">void *<b class="fsfunc">malloc</b>(</code></td><td>size_t <var class="pdparam">size</var><code>)</code>;</td></tr></table><div class="funcprototype-spacer">�</div><table border="0" summary="Function synopsis" cellspacing="0" cellpadding="0" class="funcprototype-table"><tr><td><code class="funcdef">void *<b class="fsfunc">calloc</b>(</code></td><td>size_t <var class="pdparam">number</var>, </td></tr><tr><td>�</td><td>size_t <var class="pdparam">size</var><code>)</code>;</td></tr></table><div class="funcprototype-spacer">�</div><table border="0" summary="Function synopsis" cellspacing="0" cellpadding="0" class="funcprototype-table"><tr><td><code class="funcdef">int <b class="fsfunc">posix_memalign</b>(</code></td><td>void **<var class="pdparam">ptr</var>, </td></tr><tr><td>�</td><td>size_t <var class="pdparam">alignment</var>, </td></tr><tr><td>�</td><td>size_t <var class="pdparam">size</var><code>)</code>;</td></tr></table><div class="funcprototype-spacer">�</div><table border="0" summary="Function synopsis" cellspacing="0" cellpadding="0" class="funcprototype-table"><tr><td><code class="funcdef">void *<b class="fsfunc">aligned_alloc</b>(</code></td><td>size_t <var class="pdparam">alignment</var>, </td></tr><tr><td>�</td><td>size_t <var class="pdparam">size</var><code>)</code>;</td></tr></table><div class="funcprototype-spacer">�</div><table border="0" summary="Function synopsis" cellspacing="0" cellpadding="0" class="funcprototype-table"><tr><td><code class="funcdef">void *<b class="fsfunc">realloc</b>(</code></td><td>void *<var class="pdparam">ptr</var>, </td></tr><tr><td>�</td><td>size_t <var class="pdparam">size</var><code>)</code>;</td></tr></table><div class="funcprototype-spacer">�</div><table border="0" summary="Function synopsis" cellspacing="0" cellpadding="0" class="funcprototype-table"><tr><td><code class="funcdef">void <b class="fsfunc">free</b>(</code></td><td>void *<var class="pdparam">ptr</var><code>)</code>;</td></tr></table><div class="funcprototype-spacer">�</div></div><div class="refsect2" title="Non-standard API"><a name="id286900549"></a><h3>Non-standard API</h3><table border="0" summary="Function synopsis" cellspacing="0" cellpadding="0" class="funcprototype-table"><tr><td><code class="funcdef">size_t <b class="fsfunc">malloc_usable_size</b>(</code></td><td>const void *<var class="pdparam">ptr</var><code>)</code>;</td></tr></table><div class="funcprototype-spacer">�</div><table border="0" summary="Function synopsis" cellspacing="0" cellpadding="0" class="funcprototype-table"><tr><td><code class="funcdef">void <b class="fsfunc">malloc_stats_print</b>(</code></td><td>void <var class="pdparam">(*write_cb)</var>
4            <code>(</code>void *, const char *<code>)</code>
5          , </td></tr><tr><td>�</td><td>void *<var class="pdparam">cbopaque</var>, </td></tr><tr><td>�</td><td>const char *<var class="pdparam">opts</var><code>)</code>;</td></tr></table><div class="funcprototype-spacer">�</div><table border="0" summary="Function synopsis" cellspacing="0" cellpadding="0" class="funcprototype-table"><tr><td><code class="funcdef">int <b class="fsfunc">mallctl</b>(</code></td><td>const char *<var class="pdparam">name</var>, </td></tr><tr><td>�</td><td>void *<var class="pdparam">oldp</var>, </td></tr><tr><td>�</td><td>size_t *<var class="pdparam">oldlenp</var>, </td></tr><tr><td>�</td><td>void *<var class="pdparam">newp</var>, </td></tr><tr><td>�</td><td>size_t <var class="pdparam">newlen</var><code>)</code>;</td></tr></table><div class="funcprototype-spacer">�</div><table border="0" summary="Function synopsis" cellspacing="0" cellpadding="0" class="funcprototype-table"><tr><td><code class="funcdef">int <b class="fsfunc">mallctlnametomib</b>(</code></td><td>const char *<var class="pdparam">name</var>, </td></tr><tr><td>�</td><td>size_t *<var class="pdparam">mibp</var>, </td></tr><tr><td>�</td><td>size_t *<var class="pdparam">miblenp</var><code>)</code>;</td></tr></table><div class="funcprototype-spacer">�</div><table border="0" summary="Function synopsis" cellspacing="0" cellpadding="0" class="funcprototype-table"><tr><td><code class="funcdef">int <b class="fsfunc">mallctlbymib</b>(</code></td><td>const size_t *<var class="pdparam">mib</var>, </td></tr><tr><td>�</td><td>size_t <var class="pdparam">miblen</var>, </td></tr><tr><td>�</td><td>void *<var class="pdparam">oldp</var>, </td></tr><tr><td>�</td><td>size_t *<var class="pdparam">oldlenp</var>, </td></tr><tr><td>�</td><td>void *<var class="pdparam">newp</var>, </td></tr><tr><td>�</td><td>size_t <var class="pdparam">newlen</var><code>)</code>;</td></tr></table><div class="funcprototype-spacer">�</div><table border="0" summary="Function synopsis" cellspacing="0" cellpadding="0" class="funcprototype-table"><tr><td><code class="funcdef">void <b class="fsfunc">(*malloc_message)</b>(</code></td><td>void *<var class="pdparam">cbopaque</var>, </td></tr><tr><td>�</td><td>const char *<var class="pdparam">s</var><code>)</code>;</td></tr></table><div class="funcprototype-spacer">�</div><p><span class="type">const char *</span><code class="varname">malloc_conf</code>;</p></div><div class="refsect2" title="Experimental API"><a name="id286900756"></a><h3>Experimental API</h3><table border="0" summary="Function synopsis" cellspacing="0" cellpadding="0" class="funcprototype-table"><tr><td><code class="funcdef">int <b class="fsfunc">allocm</b>(</code></td><td>void **<var class="pdparam">ptr</var>, </td></tr><tr><td>�</td><td>size_t *<var class="pdparam">rsize</var>, </td></tr><tr><td>�</td><td>size_t <var class="pdparam">size</var>, </td></tr><tr><td>�</td><td>int <var class="pdparam">flags</var><code>)</code>;</td></tr></table><div class="funcprototype-spacer">�</div><table border="0" summary="Function synopsis" cellspacing="0" cellpadding="0" class="funcprototype-table"><tr><td><code class="funcdef">int <b class="fsfunc">rallocm</b>(</code></td><td>void **<var class="pdparam">ptr</var>, </td></tr><tr><td>�</td><td>size_t *<var class="pdparam">rsize</var>, </td></tr><tr><td>�</td><td>size_t <var class="pdparam">size</var>, </td></tr><tr><td>�</td><td>size_t <var class="pdparam">extra</var>, </td></tr><tr><td>�</td><td>int <var class="pdparam">flags</var><code>)</code>;</td></tr></table><div class="funcprototype-spacer">�</div><table border="0" summary="Function synopsis" cellspacing="0" cellpadding="0" class="funcprototype-table"><tr><td><code class="funcdef">int <b class="fsfunc">sallocm</b>(</code></td><td>const void *<var class="pdparam">ptr</var>, </td></tr><tr><td>�</td><td>size_t *<var class="pdparam">rsize</var>, </td></tr><tr><td>�</td><td>int <var class="pdparam">flags</var><code>)</code>;</td></tr></table><div class="funcprototype-spacer">�</div><table border="0" summary="Function synopsis" cellspacing="0" cellpadding="0" class="funcprototype-table"><tr><td><code class="funcdef">int <b class="fsfunc">dallocm</b>(</code></td><td>void *<var class="pdparam">ptr</var>, </td></tr><tr><td>�</td><td>int <var class="pdparam">flags</var><code>)</code>;</td></tr></table><div class="funcprototype-spacer">�</div><table border="0" summary="Function synopsis" cellspacing="0" cellpadding="0" class="funcprototype-table"><tr><td><code class="funcdef">int <b class="fsfunc">nallocm</b>(</code></td><td>size_t *<var class="pdparam">rsize</var>, </td></tr><tr><td>�</td><td>size_t <var class="pdparam">size</var>, </td></tr><tr><td>�</td><td>int <var class="pdparam">flags</var><code>)</code>;</td></tr></table><div class="funcprototype-spacer">�</div></div></div></div><div class="refsect1" title="DESCRIPTION"><a name="description"></a><h2>DESCRIPTION</h2><div class="refsect2" title="Standard API"><a name="id286949297"></a><h3>Standard API</h3><p>The <code class="function">malloc</code>(<em class="parameter"><code></code></em>) function allocates
6      <em class="parameter"><code>size</code></em> bytes of uninitialized memory.  The allocated
7      space is suitably aligned (after possible pointer coercion) for storage
8      of any type of object.</p><p>The <code class="function">calloc</code>(<em class="parameter"><code></code></em>) function allocates
9      space for <em class="parameter"><code>number</code></em> objects, each
10      <em class="parameter"><code>size</code></em> bytes in length.  The result is identical to
11      calling <code class="function">malloc</code>(<em class="parameter"><code></code></em>) with an argument of
12      <em class="parameter"><code>number</code></em> * <em class="parameter"><code>size</code></em>, with the
13      exception that the allocated memory is explicitly initialized to zero
14      bytes.</p><p>The <code class="function">posix_memalign</code>(<em class="parameter"><code></code></em>) function
15      allocates <em class="parameter"><code>size</code></em> bytes of memory such that the
16      allocation's base address is an even multiple of
17      <em class="parameter"><code>alignment</code></em>, and returns the allocation in the value
18      pointed to by <em class="parameter"><code>ptr</code></em>.  The requested
19      <em class="parameter"><code>alignment</code></em> must be a power of 2 at least as large
20      as <code class="code">sizeof(<span class="type">void *</span>)</code>.</p><p>The <code class="function">aligned_alloc</code>(<em class="parameter"><code></code></em>) function
21      allocates <em class="parameter"><code>size</code></em> bytes of memory such that the
22      allocation's base address is an even multiple of
23      <em class="parameter"><code>alignment</code></em>.  The requested
24      <em class="parameter"><code>alignment</code></em> must be a power of 2.  Behavior is
25      undefined if <em class="parameter"><code>size</code></em> is not an integral multiple of
26      <em class="parameter"><code>alignment</code></em>.</p><p>The <code class="function">realloc</code>(<em class="parameter"><code></code></em>) function changes the
27      size of the previously allocated memory referenced by
28      <em class="parameter"><code>ptr</code></em> to <em class="parameter"><code>size</code></em> bytes.  The
29      contents of the memory are unchanged up to the lesser of the new and old
30      sizes.  If the new size is larger, the contents of the newly allocated
31      portion of the memory are undefined.  Upon success, the memory referenced
32      by <em class="parameter"><code>ptr</code></em> is freed and a pointer to the newly
33      allocated memory is returned.  Note that
34      <code class="function">realloc</code>(<em class="parameter"><code></code></em>) may move the memory allocation,
35      resulting in a different return value than <em class="parameter"><code>ptr</code></em>.
36      If <em class="parameter"><code>ptr</code></em> is <code class="constant">NULL</code>, the
37      <code class="function">realloc</code>(<em class="parameter"><code></code></em>) function behaves identically to
38      <code class="function">malloc</code>(<em class="parameter"><code></code></em>) for the specified size.</p><p>The <code class="function">free</code>(<em class="parameter"><code></code></em>) function causes the
39      allocated memory referenced by <em class="parameter"><code>ptr</code></em> to be made
40      available for future allocations.  If <em class="parameter"><code>ptr</code></em> is
41      <code class="constant">NULL</code>, no action occurs.</p></div><div class="refsect2" title="Non-standard API"><a name="id286949561"></a><h3>Non-standard API</h3><p>The <code class="function">malloc_usable_size</code>(<em class="parameter"><code></code></em>) function
42      returns the usable size of the allocation pointed to by
43      <em class="parameter"><code>ptr</code></em>.  The return value may be larger than the size
44      that was requested during allocation.  The
45      <code class="function">malloc_usable_size</code>(<em class="parameter"><code></code></em>) function is not a
46      mechanism for in-place <code class="function">realloc</code>(<em class="parameter"><code></code></em>); rather
47      it is provided solely as a tool for introspection purposes.  Any
48      discrepancy between the requested allocation size and the size reported
49      by <code class="function">malloc_usable_size</code>(<em class="parameter"><code></code></em>) should not be
50      depended on, since such behavior is entirely implementation-dependent.
51      </p><p>The <code class="function">malloc_stats_print</code>(<em class="parameter"><code></code></em>) function
52      writes human-readable summary statistics via the
53      <em class="parameter"><code>write_cb</code></em> callback function pointer and
54      <em class="parameter"><code>cbopaque</code></em> data passed to
55      <em class="parameter"><code>write_cb</code></em>, or
56      <code class="function">malloc_message</code>(<em class="parameter"><code></code></em>) if
57      <em class="parameter"><code>write_cb</code></em> is <code class="constant">NULL</code>.  This
58      function can be called repeatedly.  General information that never
59      changes during execution can be omitted by specifying "g" as a character
60      within the <em class="parameter"><code>opts</code></em> string.  Note that
61      <code class="function">malloc_message</code>(<em class="parameter"><code></code></em>) uses the
62      <code class="function">mallctl*</code>(<em class="parameter"><code></code></em>) functions internally, so
63      inconsistent statistics can be reported if multiple threads use these
64      functions simultaneously.  If <code class="option">--enable-stats</code> is
65      specified during configuration, &#8220;m&#8221; and &#8220;a&#8221; can
66      be specified to omit merged arena and per arena statistics, respectively;
67      &#8220;b&#8221; and &#8220;l&#8221; can be specified to omit per size
68      class statistics for bins and large objects, respectively.  Unrecognized
69      characters are silently ignored.  Note that thread caching may prevent
70      some statistics from being completely up to date, since extra locking
71      would be required to merge counters that track thread cache operations.
72      </p><p>The <code class="function">mallctl</code>(<em class="parameter"><code></code></em>) function provides a
73      general interface for introspecting the memory allocator, as well as
74      setting modifiable parameters and triggering actions.  The
75      period-separated <em class="parameter"><code>name</code></em> argument specifies a
76      location in a tree-structured namespace; see the <a class="xref" href="#mallctl_namespace" title="MALLCTL NAMESPACE">MALLCTL NAMESPACE</a> section for
77      documentation on the tree contents.  To read a value, pass a pointer via
78      <em class="parameter"><code>oldp</code></em> to adequate space to contain the value, and a
79      pointer to its length via <em class="parameter"><code>oldlenp</code></em>; otherwise pass
80      <code class="constant">NULL</code> and <code class="constant">NULL</code>.  Similarly, to
81      write a value, pass a pointer to the value via
82      <em class="parameter"><code>newp</code></em>, and its length via
83      <em class="parameter"><code>newlen</code></em>; otherwise pass <code class="constant">NULL</code>
84      and <code class="constant">0</code>.</p><p>The <code class="function">mallctlnametomib</code>(<em class="parameter"><code></code></em>) function
85      provides a way to avoid repeated name lookups for applications that
86      repeatedly query the same portion of the namespace, by translating a name
87      to a &#8220;Management Information Base&#8221; (MIB) that can be passed
88      repeatedly to <code class="function">mallctlbymib</code>(<em class="parameter"><code></code></em>).  Upon
89      successful return from <code class="function">mallctlnametomib</code>(<em class="parameter"><code></code></em>),
90      <em class="parameter"><code>mibp</code></em> contains an array of
91      <em class="parameter"><code>*miblenp</code></em> integers, where
92      <em class="parameter"><code>*miblenp</code></em> is the lesser of the number of components
93      in <em class="parameter"><code>name</code></em> and the input value of
94      <em class="parameter"><code>*miblenp</code></em>.  Thus it is possible to pass a
95      <em class="parameter"><code>*miblenp</code></em> that is smaller than the number of
96      period-separated name components, which results in a partial MIB that can
97      be used as the basis for constructing a complete MIB.  For name
98      components that are integers (e.g. the 2 in
99      <a class="link" href="#arenas.bin.i.size">
100    "<code class="mallctl">arenas.bin.2.size</code>"
101  </a>),
102      the corresponding MIB component will always be that integer.  Therefore,
103      it is legitimate to construct code like the following: </p><pre class="programlisting">
104unsigned nbins, i;
105
106int mib[4];
107size_t len, miblen;
108
109len = sizeof(nbins);
110mallctl("arenas.nbins", &amp;nbins, &amp;len, NULL, 0);
111
112miblen = 4;
113mallnametomib("arenas.bin.0.size", mib, &amp;miblen);
114for (i = 0; i &lt; nbins; i++) {
115	size_t bin_size;
116
117	mib[2] = i;
118	len = sizeof(bin_size);
119	mallctlbymib(mib, miblen, &amp;bin_size, &amp;len, NULL, 0);
120	/* Do something with bin_size... */
121}</pre></div><div class="refsect2" title="Experimental API"><a name="id286949870"></a><h3>Experimental API</h3><p>The experimental API is subject to change or removal without regard
122      for backward compatibility.  If <code class="option">--disable-experimental</code>
123      is specified during configuration, the experimental API is
124      omitted.</p><p>The <code class="function">allocm</code>(<em class="parameter"><code></code></em>),
125      <code class="function">rallocm</code>(<em class="parameter"><code></code></em>),
126      <code class="function">sallocm</code>(<em class="parameter"><code></code></em>),
127      <code class="function">dallocm</code>(<em class="parameter"><code></code></em>), and
128      <code class="function">nallocm</code>(<em class="parameter"><code></code></em>) functions all have a
129      <em class="parameter"><code>flags</code></em> argument that can be used to specify
130      options.  The functions only check the options that are contextually
131      relevant.  Use bitwise or (<code class="code">|</code>) operations to
132      specify one or more of the following:
133        </p><div class="variablelist"><dl><dt><span class="term"><code class="constant">ALLOCM_LG_ALIGN(<em class="parameter"><code>la</code></em>)
134            </code></span></dt><dd><p>Align the memory allocation to start at an address
135            that is a multiple of <code class="code">(1 &lt;&lt;
136            <em class="parameter"><code>la</code></em>)</code>.  This macro does not validate
137            that <em class="parameter"><code>la</code></em> is within the valid
138            range.</p></dd><dt><span class="term"><code class="constant">ALLOCM_ALIGN(<em class="parameter"><code>a</code></em>)
139            </code></span></dt><dd><p>Align the memory allocation to start at an address
140            that is a multiple of <em class="parameter"><code>a</code></em>, where
141            <em class="parameter"><code>a</code></em> is a power of two.  This macro does not
142            validate that <em class="parameter"><code>a</code></em> is a power of 2.
143            </p></dd><dt><span class="term"><code class="constant">ALLOCM_ZERO</code></span></dt><dd><p>Initialize newly allocated memory to contain zero
144            bytes.  In the growing reallocation case, the real size prior to
145            reallocation defines the boundary between untouched bytes and those
146            that are initialized to contain zero bytes.  If this option is
147            absent, newly allocated memory is uninitialized.</p></dd><dt><span class="term"><code class="constant">ALLOCM_NO_MOVE</code></span></dt><dd><p>For reallocation, fail rather than moving the
148            object.  This constraint can apply to both growth and
149            shrinkage.</p></dd><dt><span class="term"><code class="constant">ALLOCM_ARENA(<em class="parameter"><code>a</code></em>)
150            </code></span></dt><dd><p>Use the arena specified by the index
151            <em class="parameter"><code>a</code></em>.  This macro does not validate that
152            <em class="parameter"><code>a</code></em> specifies an arena in the valid
153            range.</p></dd></dl></div><p>
154      </p><p>The <code class="function">allocm</code>(<em class="parameter"><code></code></em>) function allocates at
155      least <em class="parameter"><code>size</code></em> bytes of memory, sets
156      <em class="parameter"><code>*ptr</code></em> to the base address of the allocation, and
157      sets <em class="parameter"><code>*rsize</code></em> to the real size of the allocation if
158      <em class="parameter"><code>rsize</code></em> is not <code class="constant">NULL</code>.  Behavior
159      is undefined if <em class="parameter"><code>size</code></em> is
160      <code class="constant">0</code>.</p><p>The <code class="function">rallocm</code>(<em class="parameter"><code></code></em>) function resizes the
161      allocation at <em class="parameter"><code>*ptr</code></em> to be at least
162      <em class="parameter"><code>size</code></em> bytes, sets <em class="parameter"><code>*ptr</code></em> to
163      the base address of the allocation if it moved, and sets
164      <em class="parameter"><code>*rsize</code></em> to the real size of the allocation if
165      <em class="parameter"><code>rsize</code></em> is not <code class="constant">NULL</code>.  If
166      <em class="parameter"><code>extra</code></em> is non-zero, an attempt is made to resize
167      the allocation to be at least <code class="code"><em class="parameter"><code>size</code></em> +
168      <em class="parameter"><code>extra</code></em>)</code> bytes, though inability to allocate
169      the extra byte(s) will not by itself result in failure.  Behavior is
170      undefined if <em class="parameter"><code>size</code></em> is <code class="constant">0</code>, or if
171      <code class="code">(<em class="parameter"><code>size</code></em> +
172      <em class="parameter"><code>extra</code></em> &gt;
173      <code class="constant">SIZE_T_MAX</code>)</code>.</p><p>The <code class="function">sallocm</code>(<em class="parameter"><code></code></em>) function sets
174      <em class="parameter"><code>*rsize</code></em> to the real size of the allocation.</p><p>The <code class="function">dallocm</code>(<em class="parameter"><code></code></em>) function causes the
175      memory referenced by <em class="parameter"><code>ptr</code></em> to be made available for
176      future allocations.</p><p>The <code class="function">nallocm</code>(<em class="parameter"><code></code></em>) function allocates no
177      memory, but it performs the same size computation as the
178      <code class="function">allocm</code>(<em class="parameter"><code></code></em>) function, and if
179      <em class="parameter"><code>rsize</code></em> is not <code class="constant">NULL</code> it sets
180      <em class="parameter"><code>*rsize</code></em> to the real size of the allocation that
181      would result from the equivalent <code class="function">allocm</code>(<em class="parameter"><code></code></em>)
182      function call.  Behavior is undefined if
183      <em class="parameter"><code>size</code></em> is <code class="constant">0</code>.</p></div></div><div class="refsect1" title="TUNING"><a name="tuning"></a><h2>TUNING</h2><p>Once, when the first call is made to one of the memory allocation
184    routines, the allocator initializes its internals based in part on various
185    options that can be specified at compile- or run-time.</p><p>The string pointed to by the global variable
186    <code class="varname">malloc_conf</code>, the &#8220;name&#8221; of the file
187    referenced by the symbolic link named <code class="filename">/etc/malloc.conf</code>, and the value of the
188    environment variable <code class="envar">MALLOC_CONF</code>, will be interpreted, in
189    that order, from left to right as options.</p><p>An options string is a comma-separated list of option:value pairs.
190    There is one key corresponding to each <a class="link" href="#opt.abort">
191    "<code class="mallctl">opt.*</code>"
192  </a> mallctl (see the <a class="xref" href="#mallctl_namespace" title="MALLCTL NAMESPACE">MALLCTL NAMESPACE</a> section for options
193    documentation).  For example, <code class="literal">abort:true,narenas:1</code> sets
194    the <a class="link" href="#opt.abort">
195    "<code class="mallctl">opt.abort</code>"
196  </a> and <a class="link" href="#opt.narenas">
197    "<code class="mallctl">opt.narenas</code>"
198  </a> options.  Some
199    options have boolean values (true/false), others have integer values (base
200    8, 10, or 16, depending on prefix), and yet others have raw string
201    values.</p></div><div class="refsect1" title="IMPLEMENTATION NOTES"><a name="implementation_notes"></a><h2>IMPLEMENTATION NOTES</h2><p>Traditionally, allocators have used
202    <span class="citerefentry"><span class="refentrytitle">sbrk</span>(2)</span> to obtain memory, which is
203    suboptimal for several reasons, including race conditions, increased
204    fragmentation, and artificial limitations on maximum usable memory.  If
205    <code class="option">--enable-dss</code> is specified during configuration, this
206    allocator uses both <span class="citerefentry"><span class="refentrytitle">mmap</span>(2)</span> and
207    <span class="citerefentry"><span class="refentrytitle">sbrk</span>(2)</span>, in that order of preference;
208    otherwise only <span class="citerefentry"><span class="refentrytitle">mmap</span>(2)</span> is used.</p><p>This allocator uses multiple arenas in order to reduce lock
209    contention for threaded programs on multi-processor systems.  This works
210    well with regard to threading scalability, but incurs some costs.  There is
211    a small fixed per-arena overhead, and additionally, arenas manage memory
212    completely independently of each other, which means a small fixed increase
213    in overall memory fragmentation.  These overheads are not generally an
214    issue, given the number of arenas normally used.  Note that using
215    substantially more arenas than the default is not likely to improve
216    performance, mainly due to reduced cache performance.  However, it may make
217    sense to reduce the number of arenas if an application does not make much
218    use of the allocation functions.</p><p>In addition to multiple arenas, unless
219    <code class="option">--disable-tcache</code> is specified during configuration, this
220    allocator supports thread-specific caching for small and large objects, in
221    order to make it possible to completely avoid synchronization for most
222    allocation requests.  Such caching allows very fast allocation in the
223    common case, but it increases memory usage and fragmentation, since a
224    bounded number of objects can remain allocated in each thread cache.</p><p>Memory is conceptually broken into equal-sized chunks, where the
225    chunk size is a power of two that is greater than the page size.  Chunks
226    are always aligned to multiples of the chunk size.  This alignment makes it
227    possible to find metadata for user objects very quickly.</p><p>User objects are broken into three categories according to size:
228    small, large, and huge.  Small objects are smaller than one page.  Large
229    objects are smaller than the chunk size.  Huge objects are a multiple of
230    the chunk size.  Small and large objects are managed by arenas; huge
231    objects are managed separately in a single data structure that is shared by
232    all threads.  Huge objects are used by applications infrequently enough
233    that this single data structure is not a scalability issue.</p><p>Each chunk that is managed by an arena tracks its contents as runs of
234    contiguous pages (unused, backing a set of small objects, or backing one
235    large object).  The combination of chunk alignment and chunk page maps
236    makes it possible to determine all metadata regarding small and large
237    allocations in constant time.</p><p>Small objects are managed in groups by page runs.  Each run maintains
238    a frontier and free list to track which regions are in use.  Allocation
239    requests that are no more than half the quantum (8 or 16, depending on
240    architecture) are rounded up to the nearest power of two that is at least
241    <code class="code">sizeof(<span class="type">double</span>)</code>.  All other small
242    object size classes are multiples of the quantum, spaced such that internal
243    fragmentation is limited to approximately 25% for all but the smallest size
244    classes.  Allocation requests that are larger than the maximum small size
245    class, but small enough to fit in an arena-managed chunk (see the <a class="link" href="#opt.lg_chunk">
246    "<code class="mallctl">opt.lg_chunk</code>"
247  </a> option), are
248    rounded up to the nearest run size.  Allocation requests that are too large
249    to fit in an arena-managed chunk are rounded up to the nearest multiple of
250    the chunk size.</p><p>Allocations are packed tightly together, which can be an issue for
251    multi-threaded applications.  If you need to assure that allocations do not
252    suffer from cacheline sharing, round your allocation requests up to the
253    nearest multiple of the cacheline size, or specify cacheline alignment when
254    allocating.</p><p>Assuming 4 MiB chunks, 4 KiB pages, and a 16-byte quantum on a 64-bit
255    system, the size classes in each category are as shown in <a class="xref" href="#size_classes" title="Table�1.�Size classes">Table 1</a>.</p><div class="table"><a name="size_classes"></a><p class="title"><b>Table�1.�Size classes</b></p><div class="table-contents"><table summary="Size classes" border="1"><colgroup><col align="left" class="c1"><col align="right" class="c2"><col align="left" class="c3"></colgroup><thead><tr><th align="left">Category</th><th align="right">Spacing</th><th align="left">Size</th></tr></thead><tbody><tr><td rowspan="7" align="left">Small</td><td align="right">lg</td><td align="left">[8]</td></tr><tr><td align="right">16</td><td align="left">[16, 32, 48, ..., 128]</td></tr><tr><td align="right">32</td><td align="left">[160, 192, 224, 256]</td></tr><tr><td align="right">64</td><td align="left">[320, 384, 448, 512]</td></tr><tr><td align="right">128</td><td align="left">[640, 768, 896, 1024]</td></tr><tr><td align="right">256</td><td align="left">[1280, 1536, 1792, 2048]</td></tr><tr><td align="right">512</td><td align="left">[2560, 3072, 3584]</td></tr><tr><td align="left">Large</td><td align="right">4 KiB</td><td align="left">[4 KiB, 8 KiB, 12 KiB, ..., 4072 KiB]</td></tr><tr><td align="left">Huge</td><td align="right">4 MiB</td><td align="left">[4 MiB, 8 MiB, 12 MiB, ...]</td></tr></tbody></table></div></div><br class="table-break"></div><div class="refsect1" title="MALLCTL NAMESPACE"><a name="mallctl_namespace"></a><h2>MALLCTL NAMESPACE</h2><p>The following names are defined in the namespace accessible via the
256    <code class="function">mallctl*</code>(<em class="parameter"><code></code></em>) functions.  Value types are
257    specified in parentheses, their readable/writable statuses are encoded as
258    <code class="literal">rw</code>, <code class="literal">r-</code>, <code class="literal">-w</code>, or
259    <code class="literal">--</code>, and required build configuration flags follow, if
260    any.  A name element encoded as <code class="literal">&lt;i&gt;</code> or
261    <code class="literal">&lt;j&gt;</code> indicates an integer component, where the
262    integer varies from 0 to some upper value that must be determined via
263    introspection.  In the case of
264    "<code class="mallctl">stats.arenas.&lt;i&gt;.*</code>"
265  ,
266    <code class="literal">&lt;i&gt;</code> equal to <a class="link" href="#arenas.narenas">
267    "<code class="mallctl">arenas.narenas</code>"
268  </a> can be
269    used to access the summation of statistics from all arenas.  Take special
270    note of the <a class="link" href="#epoch">
271    "<code class="mallctl">epoch</code>"
272  </a> mallctl,
273    which controls refreshing of cached dynamic statistics.</p><div class="variablelist"><dl><dt><span class="term">
274
275    "<code class="mallctl">version</code>"
276
277          (<span class="type">const char *</span>)
278          <code class="literal">r-</code>
279        </span></dt><dd><p>Return the jemalloc version string.</p></dd><dt><a name="epoch"></a><span class="term">
280
281    "<code class="mallctl">epoch</code>"
282
283          (<span class="type">uint64_t</span>)
284          <code class="literal">rw</code>
285        </span></dt><dd><p>If a value is passed in, refresh the data from which
286        the <code class="function">mallctl*</code>(<em class="parameter"><code></code></em>) functions report values,
287        and increment the epoch.  Return the current epoch.  This is useful for
288        detecting whether another thread caused a refresh.</p></dd><dt><span class="term">
289
290    "<code class="mallctl">config.debug</code>"
291
292          (<span class="type">bool</span>)
293          <code class="literal">r-</code>
294        </span></dt><dd><p><code class="option">--enable-debug</code> was specified during
295        build configuration.</p></dd><dt><span class="term">
296
297    "<code class="mallctl">config.dss</code>"
298
299          (<span class="type">bool</span>)
300          <code class="literal">r-</code>
301        </span></dt><dd><p><code class="option">--enable-dss</code> was specified during
302        build configuration.</p></dd><dt><span class="term">
303
304    "<code class="mallctl">config.fill</code>"
305
306          (<span class="type">bool</span>)
307          <code class="literal">r-</code>
308        </span></dt><dd><p><code class="option">--enable-fill</code> was specified during
309        build configuration.</p></dd><dt><span class="term">
310
311    "<code class="mallctl">config.lazy_lock</code>"
312
313          (<span class="type">bool</span>)
314          <code class="literal">r-</code>
315        </span></dt><dd><p><code class="option">--enable-lazy-lock</code> was specified
316        during build configuration.</p></dd><dt><span class="term">
317
318    "<code class="mallctl">config.mremap</code>"
319
320          (<span class="type">bool</span>)
321          <code class="literal">r-</code>
322        </span></dt><dd><p><code class="option">--enable-mremap</code> was specified during
323        build configuration.</p></dd><dt><span class="term">
324
325    "<code class="mallctl">config.munmap</code>"
326
327          (<span class="type">bool</span>)
328          <code class="literal">r-</code>
329        </span></dt><dd><p><code class="option">--enable-munmap</code> was specified during
330        build configuration.</p></dd><dt><span class="term">
331
332    "<code class="mallctl">config.prof</code>"
333
334          (<span class="type">bool</span>)
335          <code class="literal">r-</code>
336        </span></dt><dd><p><code class="option">--enable-prof</code> was specified during
337        build configuration.</p></dd><dt><span class="term">
338
339    "<code class="mallctl">config.prof_libgcc</code>"
340
341          (<span class="type">bool</span>)
342          <code class="literal">r-</code>
343        </span></dt><dd><p><code class="option">--disable-prof-libgcc</code> was not
344        specified during build configuration.</p></dd><dt><span class="term">
345
346    "<code class="mallctl">config.prof_libunwind</code>"
347
348          (<span class="type">bool</span>)
349          <code class="literal">r-</code>
350        </span></dt><dd><p><code class="option">--enable-prof-libunwind</code> was specified
351        during build configuration.</p></dd><dt><span class="term">
352
353    "<code class="mallctl">config.stats</code>"
354
355          (<span class="type">bool</span>)
356          <code class="literal">r-</code>
357        </span></dt><dd><p><code class="option">--enable-stats</code> was specified during
358        build configuration.</p></dd><dt><span class="term">
359
360    "<code class="mallctl">config.tcache</code>"
361
362          (<span class="type">bool</span>)
363          <code class="literal">r-</code>
364        </span></dt><dd><p><code class="option">--disable-tcache</code> was not specified
365        during build configuration.</p></dd><dt><span class="term">
366
367    "<code class="mallctl">config.tls</code>"
368
369          (<span class="type">bool</span>)
370          <code class="literal">r-</code>
371        </span></dt><dd><p><code class="option">--disable-tls</code> was not specified during
372        build configuration.</p></dd><dt><span class="term">
373
374    "<code class="mallctl">config.utrace</code>"
375
376          (<span class="type">bool</span>)
377          <code class="literal">r-</code>
378        </span></dt><dd><p><code class="option">--enable-utrace</code> was specified during
379        build configuration.</p></dd><dt><span class="term">
380
381    "<code class="mallctl">config.valgrind</code>"
382
383          (<span class="type">bool</span>)
384          <code class="literal">r-</code>
385        </span></dt><dd><p><code class="option">--enable-valgrind</code> was specified during
386        build configuration.</p></dd><dt><span class="term">
387
388    "<code class="mallctl">config.xmalloc</code>"
389
390          (<span class="type">bool</span>)
391          <code class="literal">r-</code>
392        </span></dt><dd><p><code class="option">--enable-xmalloc</code> was specified during
393        build configuration.</p></dd><dt><a name="opt.abort"></a><span class="term">
394
395    "<code class="mallctl">opt.abort</code>"
396
397          (<span class="type">bool</span>)
398          <code class="literal">r-</code>
399        </span></dt><dd><p>Abort-on-warning enabled/disabled.  If true, most
400        warnings are fatal.  The process will call
401        <span class="citerefentry"><span class="refentrytitle">abort</span>(3)</span> in these cases.  This option is
402        disabled by default unless <code class="option">--enable-debug</code> is
403        specified during configuration, in which case it is enabled by default.
404        </p></dd><dt><a name="opt.lg_chunk"></a><span class="term">
405
406    "<code class="mallctl">opt.lg_chunk</code>"
407
408          (<span class="type">size_t</span>)
409          <code class="literal">r-</code>
410        </span></dt><dd><p>Virtual memory chunk size (log base 2).  The default
411        chunk size is 4 MiB (2^22).</p></dd><dt><a name="opt.dss"></a><span class="term">
412
413    "<code class="mallctl">opt.dss</code>"
414
415          (<span class="type">const char *</span>)
416          <code class="literal">r-</code>
417        </span></dt><dd><p>dss (<span class="citerefentry"><span class="refentrytitle">sbrk</span>(2)</span>) allocation precedence as
418        related to <span class="citerefentry"><span class="refentrytitle">mmap</span>(2)</span> allocation.  The following
419        settings are supported: &#8220;disabled&#8221;, &#8220;primary&#8221;,
420        and &#8220;secondary&#8221; (default).</p></dd><dt><a name="opt.narenas"></a><span class="term">
421
422    "<code class="mallctl">opt.narenas</code>"
423
424          (<span class="type">size_t</span>)
425          <code class="literal">r-</code>
426        </span></dt><dd><p>Maximum number of arenas to use for automatic
427        multiplexing of threads and arenas.  The default is four times the
428        number of CPUs, or one if there is a single CPU.</p></dd><dt><a name="opt.lg_dirty_mult"></a><span class="term">
429
430    "<code class="mallctl">opt.lg_dirty_mult</code>"
431
432          (<span class="type">ssize_t</span>)
433          <code class="literal">r-</code>
434        </span></dt><dd><p>Per-arena minimum ratio (log base 2) of active to dirty
435        pages.  Some dirty unused pages may be allowed to accumulate, within
436        the limit set by the ratio (or one chunk worth of dirty pages,
437        whichever is greater), before informing the kernel about some of those
438        pages via <span class="citerefentry"><span class="refentrytitle">madvise</span>(2)</span> or a similar system call.  This
439        provides the kernel with sufficient information to recycle dirty pages
440        if physical memory becomes scarce and the pages remain unused.  The
441        default minimum ratio is 8:1 (2^3:1); an option value of -1 will
442        disable dirty page purging.</p></dd><dt><a name="opt.stats_print"></a><span class="term">
443
444    "<code class="mallctl">opt.stats_print</code>"
445
446          (<span class="type">bool</span>)
447          <code class="literal">r-</code>
448        </span></dt><dd><p>Enable/disable statistics printing at exit.  If
449        enabled, the <code class="function">malloc_stats_print</code>(<em class="parameter"><code></code></em>)
450        function is called at program exit via an
451        <span class="citerefentry"><span class="refentrytitle">atexit</span>(3)</span> function.  If
452        <code class="option">--enable-stats</code> is specified during configuration, this
453        has the potential to cause deadlock for a multi-threaded process that
454        exits while one or more threads are executing in the memory allocation
455        functions.  Therefore, this option should only be used with care; it is
456        primarily intended as a performance tuning aid during application
457        development.  This option is disabled by default.</p></dd><dt><a name="opt.junk"></a><span class="term">
458
459    "<code class="mallctl">opt.junk</code>"
460
461          (<span class="type">bool</span>)
462          <code class="literal">r-</code>
463          [<code class="option">--enable-fill</code>]
464        </span></dt><dd><p>Junk filling enabled/disabled.  If enabled, each byte
465        of uninitialized allocated memory will be initialized to
466        <code class="literal">0xa5</code>.  All deallocated memory will be initialized to
467        <code class="literal">0x5a</code>.  This is intended for debugging and will
468        impact performance negatively.  This option is disabled by default
469        unless <code class="option">--enable-debug</code> is specified during
470        configuration, in which case it is enabled by default unless running
471        inside <a class="ulink" href="http://valgrind.org/" target="_top">Valgrind</a>.</p></dd><dt><a name="opt.quarantine"></a><span class="term">
472
473    "<code class="mallctl">opt.quarantine</code>"
474
475          (<span class="type">size_t</span>)
476          <code class="literal">r-</code>
477          [<code class="option">--enable-fill</code>]
478        </span></dt><dd><p>Per thread quarantine size in bytes.  If non-zero, each
479        thread maintains a FIFO object quarantine that stores up to the
480        specified number of bytes of memory.  The quarantined memory is not
481        freed until it is released from quarantine, though it is immediately
482        junk-filled if the <a class="link" href="#opt.junk">
483    "<code class="mallctl">opt.junk</code>"
484  </a> option is
485        enabled.  This feature is of particular use in combination with <a class="ulink" href="http://valgrind.org/" target="_top">Valgrind</a>, which can detect attempts
486        to access quarantined objects.  This is intended for debugging and will
487        impact performance negatively.  The default quarantine size is 0 unless
488        running inside Valgrind, in which case the default is 16
489        MiB.</p></dd><dt><a name="opt.redzone"></a><span class="term">
490
491    "<code class="mallctl">opt.redzone</code>"
492
493          (<span class="type">bool</span>)
494          <code class="literal">r-</code>
495          [<code class="option">--enable-fill</code>]
496        </span></dt><dd><p>Redzones enabled/disabled.  If enabled, small
497        allocations have redzones before and after them.  Furthermore, if the
498        <a class="link" href="#opt.junk">
499    "<code class="mallctl">opt.junk</code>"
500  </a> option is
501        enabled, the redzones are checked for corruption during deallocation.
502        However, the primary intended purpose of this feature is to be used in
503        combination with <a class="ulink" href="http://valgrind.org/" target="_top">Valgrind</a>,
504        which needs redzones in order to do effective buffer overflow/underflow
505        detection.  This option is intended for debugging and will impact
506        performance negatively.  This option is disabled by
507        default unless running inside Valgrind.</p></dd><dt><a name="opt.zero"></a><span class="term">
508
509    "<code class="mallctl">opt.zero</code>"
510
511          (<span class="type">bool</span>)
512          <code class="literal">r-</code>
513          [<code class="option">--enable-fill</code>]
514        </span></dt><dd><p>Zero filling enabled/disabled.  If enabled, each byte
515        of uninitialized allocated memory will be initialized to 0.  Note that
516        this initialization only happens once for each byte, so
517        <code class="function">realloc</code>(<em class="parameter"><code></code></em>) and
518        <code class="function">rallocm</code>(<em class="parameter"><code></code></em>) calls do not zero memory that
519        was previously allocated.  This is intended for debugging and will
520        impact performance negatively.  This option is disabled by default.
521        </p></dd><dt><a name="opt.utrace"></a><span class="term">
522
523    "<code class="mallctl">opt.utrace</code>"
524
525          (<span class="type">bool</span>)
526          <code class="literal">r-</code>
527          [<code class="option">--enable-utrace</code>]
528        </span></dt><dd><p>Allocation tracing based on
529        <span class="citerefentry"><span class="refentrytitle">utrace</span>(2)</span> enabled/disabled.  This option
530        is disabled by default.</p></dd><dt><a name="opt.valgrind"></a><span class="term">
531
532    "<code class="mallctl">opt.valgrind</code>"
533
534          (<span class="type">bool</span>)
535          <code class="literal">r-</code>
536          [<code class="option">--enable-valgrind</code>]
537        </span></dt><dd><p><a class="ulink" href="http://valgrind.org/" target="_top">Valgrind</a>
538        support enabled/disabled.  This option is vestigal because jemalloc
539        auto-detects whether it is running inside Valgrind.  This option is
540        disabled by default, unless running inside Valgrind.</p></dd><dt><a name="opt.xmalloc"></a><span class="term">
541
542    "<code class="mallctl">opt.xmalloc</code>"
543
544          (<span class="type">bool</span>)
545          <code class="literal">r-</code>
546          [<code class="option">--enable-xmalloc</code>]
547        </span></dt><dd><p>Abort-on-out-of-memory enabled/disabled.  If enabled,
548        rather than returning failure for any allocation function, display a
549        diagnostic message on <code class="constant">STDERR_FILENO</code> and cause the
550        program to drop core (using
551        <span class="citerefentry"><span class="refentrytitle">abort</span>(3)</span>).  If an application is
552        designed to depend on this behavior, set the option at compile time by
553        including the following in the source code:
554        </p><pre class="programlisting">
555malloc_conf = "xmalloc:true";</pre><p>
556        This option is disabled by default.</p></dd><dt><a name="opt.tcache"></a><span class="term">
557
558    "<code class="mallctl">opt.tcache</code>"
559
560          (<span class="type">bool</span>)
561          <code class="literal">r-</code>
562          [<code class="option">--enable-tcache</code>]
563        </span></dt><dd><p>Thread-specific caching enabled/disabled.  When there
564        are multiple threads, each thread uses a thread-specific cache for
565        objects up to a certain size.  Thread-specific caching allows many
566        allocations to be satisfied without performing any thread
567        synchronization, at the cost of increased memory use.  See the
568        <a class="link" href="#opt.lg_tcache_max">
569    "<code class="mallctl">opt.lg_tcache_max</code>"
570  </a>
571        option for related tuning information.  This option is enabled by
572        default unless running inside <a class="ulink" href="http://valgrind.org/" target="_top">Valgrind</a>.</p></dd><dt><a name="opt.lg_tcache_max"></a><span class="term">
573
574    "<code class="mallctl">opt.lg_tcache_max</code>"
575
576          (<span class="type">size_t</span>)
577          <code class="literal">r-</code>
578          [<code class="option">--enable-tcache</code>]
579        </span></dt><dd><p>Maximum size class (log base 2) to cache in the
580        thread-specific cache.  At a minimum, all small size classes are
581        cached, and at a maximum all large size classes are cached.  The
582        default maximum is 32 KiB (2^15).</p></dd><dt><a name="opt.prof"></a><span class="term">
583
584    "<code class="mallctl">opt.prof</code>"
585
586          (<span class="type">bool</span>)
587          <code class="literal">r-</code>
588          [<code class="option">--enable-prof</code>]
589        </span></dt><dd><p>Memory profiling enabled/disabled.  If enabled, profile
590        memory allocation activity.  See the <a class="link" href="#opt.prof_active">
591    "<code class="mallctl">opt.prof_active</code>"
592  </a>
593        option for on-the-fly activation/deactivation.  See the <a class="link" href="#opt.lg_prof_sample">
594    "<code class="mallctl">opt.lg_prof_sample</code>"
595  </a>
596        option for probabilistic sampling control.  See the <a class="link" href="#opt.prof_accum">
597    "<code class="mallctl">opt.prof_accum</code>"
598  </a>
599        option for control of cumulative sample reporting.  See the <a class="link" href="#opt.lg_prof_interval">
600    "<code class="mallctl">opt.lg_prof_interval</code>"
601  </a>
602        option for information on interval-triggered profile dumping, the <a class="link" href="#opt.prof_gdump">
603    "<code class="mallctl">opt.prof_gdump</code>"
604  </a>
605        option for information on high-water-triggered profile dumping, and the
606        <a class="link" href="#opt.prof_final">
607    "<code class="mallctl">opt.prof_final</code>"
608  </a>
609        option for final profile dumping.  Profile output is compatible with
610        the included <span class="command"><strong>pprof</strong></span> Perl script, which originates
611        from the <a class="ulink" href="http://code.google.com/p/gperftools/" target="_top">gperftools
612        package</a>.</p></dd><dt><a name="opt.prof_prefix"></a><span class="term">
613
614    "<code class="mallctl">opt.prof_prefix</code>"
615
616          (<span class="type">const char *</span>)
617          <code class="literal">r-</code>
618          [<code class="option">--enable-prof</code>]
619        </span></dt><dd><p>Filename prefix for profile dumps.  If the prefix is
620        set to the empty string, no automatic dumps will occur; this is
621        primarily useful for disabling the automatic final heap dump (which
622        also disables leak reporting, if enabled).  The default prefix is
623        <code class="filename">jeprof</code>.</p></dd><dt><a name="opt.prof_active"></a><span class="term">
624
625    "<code class="mallctl">opt.prof_active</code>"
626
627          (<span class="type">bool</span>)
628          <code class="literal">r-</code>
629          [<code class="option">--enable-prof</code>]
630        </span></dt><dd><p>Profiling activated/deactivated.  This is a secondary
631        control mechanism that makes it possible to start the application with
632        profiling enabled (see the <a class="link" href="#opt.prof">
633    "<code class="mallctl">opt.prof</code>"
634  </a> option) but
635        inactive, then toggle profiling at any time during program execution
636        with the <a class="link" href="#prof.active">
637    "<code class="mallctl">prof.active</code>"
638  </a> mallctl.
639        This option is enabled by default.</p></dd><dt><a name="opt.lg_prof_sample"></a><span class="term">
640
641    "<code class="mallctl">opt.lg_prof_sample</code>"
642
643          (<span class="type">ssize_t</span>)
644          <code class="literal">r-</code>
645          [<code class="option">--enable-prof</code>]
646        </span></dt><dd><p>Average interval (log base 2) between allocation
647        samples, as measured in bytes of allocation activity.  Increasing the
648        sampling interval decreases profile fidelity, but also decreases the
649        computational overhead.  The default sample interval is 512 KiB (2^19
650        B).</p></dd><dt><a name="opt.prof_accum"></a><span class="term">
651
652    "<code class="mallctl">opt.prof_accum</code>"
653
654          (<span class="type">bool</span>)
655          <code class="literal">r-</code>
656          [<code class="option">--enable-prof</code>]
657        </span></dt><dd><p>Reporting of cumulative object/byte counts in profile
658        dumps enabled/disabled.  If this option is enabled, every unique
659        backtrace must be stored for the duration of execution.  Depending on
660        the application, this can impose a large memory overhead, and the
661        cumulative counts are not always of interest.  This option is disabled
662        by default.</p></dd><dt><a name="opt.lg_prof_interval"></a><span class="term">
663
664    "<code class="mallctl">opt.lg_prof_interval</code>"
665
666          (<span class="type">ssize_t</span>)
667          <code class="literal">r-</code>
668          [<code class="option">--enable-prof</code>]
669        </span></dt><dd><p>Average interval (log base 2) between memory profile
670        dumps, as measured in bytes of allocation activity.  The actual
671        interval between dumps may be sporadic because decentralized allocation
672        counters are used to avoid synchronization bottlenecks.  Profiles are
673        dumped to files named according to the pattern
674        <code class="filename">&lt;prefix&gt;.&lt;pid&gt;.&lt;seq&gt;.i&lt;iseq&gt;.heap</code>,
675        where <code class="literal">&lt;prefix&gt;</code> is controlled by the
676        <a class="link" href="#opt.prof_prefix">
677    "<code class="mallctl">opt.prof_prefix</code>"
678  </a>
679        option.  By default, interval-triggered profile dumping is disabled
680        (encoded as -1).
681        </p></dd><dt><a name="opt.prof_gdump"></a><span class="term">
682
683    "<code class="mallctl">opt.prof_gdump</code>"
684
685          (<span class="type">bool</span>)
686          <code class="literal">r-</code>
687          [<code class="option">--enable-prof</code>]
688        </span></dt><dd><p>Trigger a memory profile dump every time the total
689        virtual memory exceeds the previous maximum.  Profiles are dumped to
690        files named according to the pattern
691        <code class="filename">&lt;prefix&gt;.&lt;pid&gt;.&lt;seq&gt;.u&lt;useq&gt;.heap</code>,
692        where <code class="literal">&lt;prefix&gt;</code> is controlled by the <a class="link" href="#opt.prof_prefix">
693    "<code class="mallctl">opt.prof_prefix</code>"
694  </a>
695        option.  This option is disabled by default.</p></dd><dt><a name="opt.prof_final"></a><span class="term">
696
697    "<code class="mallctl">opt.prof_final</code>"
698
699          (<span class="type">bool</span>)
700          <code class="literal">r-</code>
701          [<code class="option">--enable-prof</code>]
702        </span></dt><dd><p>Use an
703        <span class="citerefentry"><span class="refentrytitle">atexit</span>(3)</span> function to dump final memory
704        usage to a file named according to the pattern
705        <code class="filename">&lt;prefix&gt;.&lt;pid&gt;.&lt;seq&gt;.f.heap</code>,
706        where <code class="literal">&lt;prefix&gt;</code> is controlled by the <a class="link" href="#opt.prof_prefix">
707    "<code class="mallctl">opt.prof_prefix</code>"
708  </a>
709        option.  This option is enabled by default.</p></dd><dt><a name="opt.prof_leak"></a><span class="term">
710
711    "<code class="mallctl">opt.prof_leak</code>"
712
713          (<span class="type">bool</span>)
714          <code class="literal">r-</code>
715          [<code class="option">--enable-prof</code>]
716        </span></dt><dd><p>Leak reporting enabled/disabled.  If enabled, use an
717        <span class="citerefentry"><span class="refentrytitle">atexit</span>(3)</span> function to report memory leaks
718        detected by allocation sampling.  See the
719        <a class="link" href="#opt.prof">
720    "<code class="mallctl">opt.prof</code>"
721  </a> option for
722        information on analyzing heap profile output.  This option is disabled
723        by default.</p></dd><dt><span class="term">
724
725    "<code class="mallctl">thread.arena</code>"
726
727          (<span class="type">unsigned</span>)
728          <code class="literal">rw</code>
729        </span></dt><dd><p>Get or set the arena associated with the calling
730        thread.  If the specified arena was not initialized beforehand (see the
731        <a class="link" href="#arenas.initialized">
732    "<code class="mallctl">arenas.initialized</code>"
733  </a>
734        mallctl), it will be automatically initialized as a side effect of
735        calling this interface.</p></dd><dt><a name="thread.allocated"></a><span class="term">
736
737    "<code class="mallctl">thread.allocated</code>"
738
739          (<span class="type">uint64_t</span>)
740          <code class="literal">r-</code>
741          [<code class="option">--enable-stats</code>]
742        </span></dt><dd><p>Get the total number of bytes ever allocated by the
743        calling thread.  This counter has the potential to wrap around; it is
744        up to the application to appropriately interpret the counter in such
745        cases.</p></dd><dt><span class="term">
746
747    "<code class="mallctl">thread.allocatedp</code>"
748
749          (<span class="type">uint64_t *</span>)
750          <code class="literal">r-</code>
751          [<code class="option">--enable-stats</code>]
752        </span></dt><dd><p>Get a pointer to the the value that is returned by the
753        <a class="link" href="#thread.allocated">
754    "<code class="mallctl">thread.allocated</code>"
755  </a>
756        mallctl.  This is useful for avoiding the overhead of repeated
757        <code class="function">mallctl*</code>(<em class="parameter"><code></code></em>) calls.</p></dd><dt><a name="thread.deallocated"></a><span class="term">
758
759    "<code class="mallctl">thread.deallocated</code>"
760
761          (<span class="type">uint64_t</span>)
762          <code class="literal">r-</code>
763          [<code class="option">--enable-stats</code>]
764        </span></dt><dd><p>Get the total number of bytes ever deallocated by the
765        calling thread.  This counter has the potential to wrap around; it is
766        up to the application to appropriately interpret the counter in such
767        cases.</p></dd><dt><span class="term">
768
769    "<code class="mallctl">thread.deallocatedp</code>"
770
771          (<span class="type">uint64_t *</span>)
772          <code class="literal">r-</code>
773          [<code class="option">--enable-stats</code>]
774        </span></dt><dd><p>Get a pointer to the the value that is returned by the
775        <a class="link" href="#thread.deallocated">
776    "<code class="mallctl">thread.deallocated</code>"
777  </a>
778        mallctl.  This is useful for avoiding the overhead of repeated
779        <code class="function">mallctl*</code>(<em class="parameter"><code></code></em>) calls.</p></dd><dt><span class="term">
780
781    "<code class="mallctl">thread.tcache.enabled</code>"
782
783          (<span class="type">bool</span>)
784          <code class="literal">rw</code>
785          [<code class="option">--enable-tcache</code>]
786        </span></dt><dd><p>Enable/disable calling thread's tcache.  The tcache is
787        implicitly flushed as a side effect of becoming
788        disabled (see
789    "<code class="mallctl">thread.tcache.flush</code>"
790  ).
791        </p></dd><dt><span class="term">
792
793    "<code class="mallctl">thread.tcache.flush</code>"
794
795          (<span class="type">void</span>)
796          <code class="literal">--</code>
797          [<code class="option">--enable-tcache</code>]
798        </span></dt><dd><p>Flush calling thread's tcache.  This interface releases
799        all cached objects and internal data structures associated with the
800        calling thread's thread-specific cache.  Ordinarily, this interface
801        need not be called, since automatic periodic incremental garbage
802        collection occurs, and the thread cache is automatically discarded when
803        a thread exits.  However, garbage collection is triggered by allocation
804        activity, so it is possible for a thread that stops
805        allocating/deallocating to retain its cache indefinitely, in which case
806        the developer may find manual flushing useful.</p></dd><dt><a name="arena.i.purge"></a><span class="term">
807
808    "<code class="mallctl">arena.&lt;i&gt;.purge</code>"
809
810          (<span class="type">unsigned</span>)
811          <code class="literal">--</code>
812        </span></dt><dd><p>Purge unused dirty pages for arena &lt;i&gt;, or for
813        all arenas if &lt;i&gt; equals <a class="link" href="#arenas.narenas">
814    "<code class="mallctl">arenas.narenas</code>"
815  </a>.
816        </p></dd><dt><a name="arena.i.dss"></a><span class="term">
817
818    "<code class="mallctl">arena.&lt;i&gt;.dss</code>"
819
820          (<span class="type">const char *</span>)
821          <code class="literal">rw</code>
822        </span></dt><dd><p>Set the precedence of dss allocation as related to mmap
823        allocation for arena &lt;i&gt;, or for all arenas if &lt;i&gt; equals
824        <a class="link" href="#arenas.narenas">
825    "<code class="mallctl">arenas.narenas</code>"
826  </a>.  See
827        <a class="link" href="#opt.dss">
828    "<code class="mallctl">opt.dss</code>"
829  </a> for supported
830        settings.
831        </p></dd><dt><a name="arenas.narenas"></a><span class="term">
832
833    "<code class="mallctl">arenas.narenas</code>"
834
835          (<span class="type">unsigned</span>)
836          <code class="literal">r-</code>
837        </span></dt><dd><p>Current limit on number of arenas.</p></dd><dt><a name="arenas.initialized"></a><span class="term">
838
839    "<code class="mallctl">arenas.initialized</code>"
840
841          (<span class="type">bool *</span>)
842          <code class="literal">r-</code>
843        </span></dt><dd><p>An array of <a class="link" href="#arenas.narenas">
844    "<code class="mallctl">arenas.narenas</code>"
845  </a>
846        booleans.  Each boolean indicates whether the corresponding arena is
847        initialized.</p></dd><dt><span class="term">
848
849    "<code class="mallctl">arenas.quantum</code>"
850
851          (<span class="type">size_t</span>)
852          <code class="literal">r-</code>
853        </span></dt><dd><p>Quantum size.</p></dd><dt><span class="term">
854
855    "<code class="mallctl">arenas.page</code>"
856
857          (<span class="type">size_t</span>)
858          <code class="literal">r-</code>
859        </span></dt><dd><p>Page size.</p></dd><dt><span class="term">
860
861    "<code class="mallctl">arenas.tcache_max</code>"
862
863          (<span class="type">size_t</span>)
864          <code class="literal">r-</code>
865          [<code class="option">--enable-tcache</code>]
866        </span></dt><dd><p>Maximum thread-cached size class.</p></dd><dt><span class="term">
867
868    "<code class="mallctl">arenas.nbins</code>"
869
870          (<span class="type">unsigned</span>)
871          <code class="literal">r-</code>
872        </span></dt><dd><p>Number of bin size classes.</p></dd><dt><span class="term">
873
874    "<code class="mallctl">arenas.nhbins</code>"
875
876          (<span class="type">unsigned</span>)
877          <code class="literal">r-</code>
878          [<code class="option">--enable-tcache</code>]
879        </span></dt><dd><p>Total number of thread cache bin size
880        classes.</p></dd><dt><a name="arenas.bin.i.size"></a><span class="term">
881
882    "<code class="mallctl">arenas.bin.&lt;i&gt;.size</code>"
883
884          (<span class="type">size_t</span>)
885          <code class="literal">r-</code>
886        </span></dt><dd><p>Maximum size supported by size class.</p></dd><dt><span class="term">
887
888    "<code class="mallctl">arenas.bin.&lt;i&gt;.nregs</code>"
889
890          (<span class="type">uint32_t</span>)
891          <code class="literal">r-</code>
892        </span></dt><dd><p>Number of regions per page run.</p></dd><dt><span class="term">
893
894    "<code class="mallctl">arenas.bin.&lt;i&gt;.run_size</code>"
895
896          (<span class="type">size_t</span>)
897          <code class="literal">r-</code>
898        </span></dt><dd><p>Number of bytes per page run.</p></dd><dt><span class="term">
899
900    "<code class="mallctl">arenas.nlruns</code>"
901
902          (<span class="type">size_t</span>)
903          <code class="literal">r-</code>
904        </span></dt><dd><p>Total number of large size classes.</p></dd><dt><span class="term">
905
906    "<code class="mallctl">arenas.lrun.&lt;i&gt;.size</code>"
907
908          (<span class="type">size_t</span>)
909          <code class="literal">r-</code>
910        </span></dt><dd><p>Maximum size supported by this large size
911        class.</p></dd><dt><span class="term">
912
913    "<code class="mallctl">arenas.purge</code>"
914
915          (<span class="type">unsigned</span>)
916          <code class="literal">-w</code>
917        </span></dt><dd><p>Purge unused dirty pages for the specified arena, or
918        for all arenas if none is specified.</p></dd><dt><span class="term">
919
920    "<code class="mallctl">arenas.extend</code>"
921
922          (<span class="type">unsigned</span>)
923          <code class="literal">r-</code>
924        </span></dt><dd><p>Extend the array of arenas by appending a new arena,
925        and returning the new arena index.</p></dd><dt><a name="prof.active"></a><span class="term">
926
927    "<code class="mallctl">prof.active</code>"
928
929          (<span class="type">bool</span>)
930          <code class="literal">rw</code>
931          [<code class="option">--enable-prof</code>]
932        </span></dt><dd><p>Control whether sampling is currently active.  See the
933        <a class="link" href="#opt.prof_active">
934    "<code class="mallctl">opt.prof_active</code>"
935  </a>
936        option for additional information.
937        </p></dd><dt><span class="term">
938
939    "<code class="mallctl">prof.dump</code>"
940
941          (<span class="type">const char *</span>)
942          <code class="literal">-w</code>
943          [<code class="option">--enable-prof</code>]
944        </span></dt><dd><p>Dump a memory profile to the specified file, or if NULL
945        is specified, to a file according to the pattern
946        <code class="filename">&lt;prefix&gt;.&lt;pid&gt;.&lt;seq&gt;.m&lt;mseq&gt;.heap</code>,
947        where <code class="literal">&lt;prefix&gt;</code> is controlled by the
948        <a class="link" href="#opt.prof_prefix">
949    "<code class="mallctl">opt.prof_prefix</code>"
950  </a>
951        option.</p></dd><dt><span class="term">
952
953    "<code class="mallctl">prof.interval</code>"
954
955          (<span class="type">uint64_t</span>)
956          <code class="literal">r-</code>
957          [<code class="option">--enable-prof</code>]
958        </span></dt><dd><p>Average number of bytes allocated between
959        inverval-based profile dumps.  See the
960        <a class="link" href="#opt.lg_prof_interval">
961    "<code class="mallctl">opt.lg_prof_interval</code>"
962  </a>
963        option for additional information.</p></dd><dt><a name="stats.cactive"></a><span class="term">
964
965    "<code class="mallctl">stats.cactive</code>"
966
967          (<span class="type">size_t *</span>)
968          <code class="literal">r-</code>
969          [<code class="option">--enable-stats</code>]
970        </span></dt><dd><p>Pointer to a counter that contains an approximate count
971        of the current number of bytes in active pages.  The estimate may be
972        high, but never low, because each arena rounds up to the nearest
973        multiple of the chunk size when computing its contribution to the
974        counter.  Note that the <a class="link" href="#epoch">
975    "<code class="mallctl">epoch</code>"
976  </a> mallctl has no bearing
977        on this counter.  Furthermore, counter consistency is maintained via
978        atomic operations, so it is necessary to use an atomic operation in
979        order to guarantee a consistent read when dereferencing the pointer.
980        </p></dd><dt><a name="stats.allocated"></a><span class="term">
981
982    "<code class="mallctl">stats.allocated</code>"
983
984          (<span class="type">size_t</span>)
985          <code class="literal">r-</code>
986          [<code class="option">--enable-stats</code>]
987        </span></dt><dd><p>Total number of bytes allocated by the
988        application.</p></dd><dt><a name="stats.active"></a><span class="term">
989
990    "<code class="mallctl">stats.active</code>"
991
992          (<span class="type">size_t</span>)
993          <code class="literal">r-</code>
994          [<code class="option">--enable-stats</code>]
995        </span></dt><dd><p>Total number of bytes in active pages allocated by the
996        application.  This is a multiple of the page size, and greater than or
997        equal to <a class="link" href="#stats.allocated">
998    "<code class="mallctl">stats.allocated</code>"
999  </a>.
1000        This does not include <a class="link" href="#stats.arenas.i.pdirty">
1001
1002    "<code class="mallctl">stats.arenas.&lt;i&gt;.pdirty</code>"
1003  </a> and pages
1004        entirely devoted to allocator metadata.</p></dd><dt><span class="term">
1005
1006    "<code class="mallctl">stats.mapped</code>"
1007
1008          (<span class="type">size_t</span>)
1009          <code class="literal">r-</code>
1010          [<code class="option">--enable-stats</code>]
1011        </span></dt><dd><p>Total number of bytes in chunks mapped on behalf of the
1012        application.  This is a multiple of the chunk size, and is at least as
1013        large as <a class="link" href="#stats.active">
1014    "<code class="mallctl">stats.active</code>"
1015  </a>.  This
1016        does not include inactive chunks.</p></dd><dt><span class="term">
1017
1018    "<code class="mallctl">stats.chunks.current</code>"
1019
1020          (<span class="type">size_t</span>)
1021          <code class="literal">r-</code>
1022          [<code class="option">--enable-stats</code>]
1023        </span></dt><dd><p>Total number of chunks actively mapped on behalf of the
1024        application.  This does not include inactive chunks.
1025        </p></dd><dt><span class="term">
1026
1027    "<code class="mallctl">stats.chunks.total</code>"
1028
1029          (<span class="type">uint64_t</span>)
1030          <code class="literal">r-</code>
1031          [<code class="option">--enable-stats</code>]
1032        </span></dt><dd><p>Cumulative number of chunks allocated.</p></dd><dt><span class="term">
1033
1034    "<code class="mallctl">stats.chunks.high</code>"
1035
1036          (<span class="type">size_t</span>)
1037          <code class="literal">r-</code>
1038          [<code class="option">--enable-stats</code>]
1039        </span></dt><dd><p>Maximum number of active chunks at any time thus far.
1040        </p></dd><dt><span class="term">
1041
1042    "<code class="mallctl">stats.huge.allocated</code>"
1043
1044          (<span class="type">size_t</span>)
1045          <code class="literal">r-</code>
1046          [<code class="option">--enable-stats</code>]
1047        </span></dt><dd><p>Number of bytes currently allocated by huge objects.
1048        </p></dd><dt><span class="term">
1049
1050    "<code class="mallctl">stats.huge.nmalloc</code>"
1051
1052          (<span class="type">uint64_t</span>)
1053          <code class="literal">r-</code>
1054          [<code class="option">--enable-stats</code>]
1055        </span></dt><dd><p>Cumulative number of huge allocation requests.
1056        </p></dd><dt><span class="term">
1057
1058    "<code class="mallctl">stats.huge.ndalloc</code>"
1059
1060          (<span class="type">uint64_t</span>)
1061          <code class="literal">r-</code>
1062          [<code class="option">--enable-stats</code>]
1063        </span></dt><dd><p>Cumulative number of huge deallocation requests.
1064        </p></dd><dt><span class="term">
1065
1066    "<code class="mallctl">stats.arenas.&lt;i&gt;.dss</code>"
1067
1068          (<span class="type">const char *</span>)
1069          <code class="literal">r-</code>
1070        </span></dt><dd><p>dss (<span class="citerefentry"><span class="refentrytitle">sbrk</span>(2)</span>) allocation precedence as
1071        related to <span class="citerefentry"><span class="refentrytitle">mmap</span>(2)</span> allocation.  See <a class="link" href="#opt.dss">
1072    "<code class="mallctl">opt.dss</code>"
1073  </a> for details.
1074        </p></dd><dt><span class="term">
1075
1076    "<code class="mallctl">stats.arenas.&lt;i&gt;.nthreads</code>"
1077
1078          (<span class="type">unsigned</span>)
1079          <code class="literal">r-</code>
1080        </span></dt><dd><p>Number of threads currently assigned to
1081        arena.</p></dd><dt><span class="term">
1082
1083    "<code class="mallctl">stats.arenas.&lt;i&gt;.pactive</code>"
1084
1085          (<span class="type">size_t</span>)
1086          <code class="literal">r-</code>
1087        </span></dt><dd><p>Number of pages in active runs.</p></dd><dt><a name="stats.arenas.i.pdirty"></a><span class="term">
1088
1089    "<code class="mallctl">stats.arenas.&lt;i&gt;.pdirty</code>"
1090
1091          (<span class="type">size_t</span>)
1092          <code class="literal">r-</code>
1093        </span></dt><dd><p>Number of pages within unused runs that are potentially
1094        dirty, and for which <code class="function">madvise</code>(<em class="parameter"><code>...</code></em>,
1095        <em class="parameter"><code><code class="constant">MADV_DONTNEED</code></code></em>) or
1096        similar has not been called.</p></dd><dt><span class="term">
1097
1098    "<code class="mallctl">stats.arenas.&lt;i&gt;.mapped</code>"
1099
1100          (<span class="type">size_t</span>)
1101          <code class="literal">r-</code>
1102          [<code class="option">--enable-stats</code>]
1103        </span></dt><dd><p>Number of mapped bytes.</p></dd><dt><span class="term">
1104
1105    "<code class="mallctl">stats.arenas.&lt;i&gt;.npurge</code>"
1106
1107          (<span class="type">uint64_t</span>)
1108          <code class="literal">r-</code>
1109          [<code class="option">--enable-stats</code>]
1110        </span></dt><dd><p>Number of dirty page purge sweeps performed.
1111        </p></dd><dt><span class="term">
1112
1113    "<code class="mallctl">stats.arenas.&lt;i&gt;.nmadvise</code>"
1114
1115          (<span class="type">uint64_t</span>)
1116          <code class="literal">r-</code>
1117          [<code class="option">--enable-stats</code>]
1118        </span></dt><dd><p>Number of <code class="function">madvise</code>(<em class="parameter"><code>...</code></em>,
1119        <em class="parameter"><code><code class="constant">MADV_DONTNEED</code></code></em>) or
1120        similar calls made to purge dirty pages.</p></dd><dt><span class="term">
1121
1122    "<code class="mallctl">stats.arenas.&lt;i&gt;.npurged</code>"
1123
1124          (<span class="type">uint64_t</span>)
1125          <code class="literal">r-</code>
1126          [<code class="option">--enable-stats</code>]
1127        </span></dt><dd><p>Number of pages purged.</p></dd><dt><span class="term">
1128
1129    "<code class="mallctl">stats.arenas.&lt;i&gt;.small.allocated</code>"
1130
1131          (<span class="type">size_t</span>)
1132          <code class="literal">r-</code>
1133          [<code class="option">--enable-stats</code>]
1134        </span></dt><dd><p>Number of bytes currently allocated by small objects.
1135        </p></dd><dt><span class="term">
1136
1137    "<code class="mallctl">stats.arenas.&lt;i&gt;.small.nmalloc</code>"
1138
1139          (<span class="type">uint64_t</span>)
1140          <code class="literal">r-</code>
1141          [<code class="option">--enable-stats</code>]
1142        </span></dt><dd><p>Cumulative number of allocation requests served by
1143        small bins.</p></dd><dt><span class="term">
1144
1145    "<code class="mallctl">stats.arenas.&lt;i&gt;.small.ndalloc</code>"
1146
1147          (<span class="type">uint64_t</span>)
1148          <code class="literal">r-</code>
1149          [<code class="option">--enable-stats</code>]
1150        </span></dt><dd><p>Cumulative number of small objects returned to bins.
1151        </p></dd><dt><span class="term">
1152
1153    "<code class="mallctl">stats.arenas.&lt;i&gt;.small.nrequests</code>"
1154
1155          (<span class="type">uint64_t</span>)
1156          <code class="literal">r-</code>
1157          [<code class="option">--enable-stats</code>]
1158        </span></dt><dd><p>Cumulative number of small allocation requests.
1159        </p></dd><dt><span class="term">
1160
1161    "<code class="mallctl">stats.arenas.&lt;i&gt;.large.allocated</code>"
1162
1163          (<span class="type">size_t</span>)
1164          <code class="literal">r-</code>
1165          [<code class="option">--enable-stats</code>]
1166        </span></dt><dd><p>Number of bytes currently allocated by large objects.
1167        </p></dd><dt><span class="term">
1168
1169    "<code class="mallctl">stats.arenas.&lt;i&gt;.large.nmalloc</code>"
1170
1171          (<span class="type">uint64_t</span>)
1172          <code class="literal">r-</code>
1173          [<code class="option">--enable-stats</code>]
1174        </span></dt><dd><p>Cumulative number of large allocation requests served
1175        directly by the arena.</p></dd><dt><span class="term">
1176
1177    "<code class="mallctl">stats.arenas.&lt;i&gt;.large.ndalloc</code>"
1178
1179          (<span class="type">uint64_t</span>)
1180          <code class="literal">r-</code>
1181          [<code class="option">--enable-stats</code>]
1182        </span></dt><dd><p>Cumulative number of large deallocation requests served
1183        directly by the arena.</p></dd><dt><span class="term">
1184
1185    "<code class="mallctl">stats.arenas.&lt;i&gt;.large.nrequests</code>"
1186
1187          (<span class="type">uint64_t</span>)
1188          <code class="literal">r-</code>
1189          [<code class="option">--enable-stats</code>]
1190        </span></dt><dd><p>Cumulative number of large allocation requests.
1191        </p></dd><dt><span class="term">
1192
1193    "<code class="mallctl">stats.arenas.&lt;i&gt;.bins.&lt;j&gt;.allocated</code>"
1194
1195          (<span class="type">size_t</span>)
1196          <code class="literal">r-</code>
1197          [<code class="option">--enable-stats</code>]
1198        </span></dt><dd><p>Current number of bytes allocated by
1199        bin.</p></dd><dt><span class="term">
1200
1201    "<code class="mallctl">stats.arenas.&lt;i&gt;.bins.&lt;j&gt;.nmalloc</code>"
1202
1203          (<span class="type">uint64_t</span>)
1204          <code class="literal">r-</code>
1205          [<code class="option">--enable-stats</code>]
1206        </span></dt><dd><p>Cumulative number of allocations served by bin.
1207        </p></dd><dt><span class="term">
1208
1209    "<code class="mallctl">stats.arenas.&lt;i&gt;.bins.&lt;j&gt;.ndalloc</code>"
1210
1211          (<span class="type">uint64_t</span>)
1212          <code class="literal">r-</code>
1213          [<code class="option">--enable-stats</code>]
1214        </span></dt><dd><p>Cumulative number of allocations returned to bin.
1215        </p></dd><dt><span class="term">
1216
1217    "<code class="mallctl">stats.arenas.&lt;i&gt;.bins.&lt;j&gt;.nrequests</code>"
1218
1219          (<span class="type">uint64_t</span>)
1220          <code class="literal">r-</code>
1221          [<code class="option">--enable-stats</code>]
1222        </span></dt><dd><p>Cumulative number of allocation
1223        requests.</p></dd><dt><span class="term">
1224
1225    "<code class="mallctl">stats.arenas.&lt;i&gt;.bins.&lt;j&gt;.nfills</code>"
1226
1227          (<span class="type">uint64_t</span>)
1228          <code class="literal">r-</code>
1229          [<code class="option">--enable-stats</code> <code class="option">--enable-tcache</code>]
1230        </span></dt><dd><p>Cumulative number of tcache fills.</p></dd><dt><span class="term">
1231
1232    "<code class="mallctl">stats.arenas.&lt;i&gt;.bins.&lt;j&gt;.nflushes</code>"
1233
1234          (<span class="type">uint64_t</span>)
1235          <code class="literal">r-</code>
1236          [<code class="option">--enable-stats</code> <code class="option">--enable-tcache</code>]
1237        </span></dt><dd><p>Cumulative number of tcache flushes.</p></dd><dt><span class="term">
1238
1239    "<code class="mallctl">stats.arenas.&lt;i&gt;.bins.&lt;j&gt;.nruns</code>"
1240
1241          (<span class="type">uint64_t</span>)
1242          <code class="literal">r-</code>
1243          [<code class="option">--enable-stats</code>]
1244        </span></dt><dd><p>Cumulative number of runs created.</p></dd><dt><span class="term">
1245
1246    "<code class="mallctl">stats.arenas.&lt;i&gt;.bins.&lt;j&gt;.nreruns</code>"
1247
1248          (<span class="type">uint64_t</span>)
1249          <code class="literal">r-</code>
1250          [<code class="option">--enable-stats</code>]
1251        </span></dt><dd><p>Cumulative number of times the current run from which
1252        to allocate changed.</p></dd><dt><span class="term">
1253
1254    "<code class="mallctl">stats.arenas.&lt;i&gt;.bins.&lt;j&gt;.curruns</code>"
1255
1256          (<span class="type">size_t</span>)
1257          <code class="literal">r-</code>
1258          [<code class="option">--enable-stats</code>]
1259        </span></dt><dd><p>Current number of runs.</p></dd><dt><span class="term">
1260
1261    "<code class="mallctl">stats.arenas.&lt;i&gt;.lruns.&lt;j&gt;.nmalloc</code>"
1262
1263          (<span class="type">uint64_t</span>)
1264          <code class="literal">r-</code>
1265          [<code class="option">--enable-stats</code>]
1266        </span></dt><dd><p>Cumulative number of allocation requests for this size
1267        class served directly by the arena.</p></dd><dt><span class="term">
1268
1269    "<code class="mallctl">stats.arenas.&lt;i&gt;.lruns.&lt;j&gt;.ndalloc</code>"
1270
1271          (<span class="type">uint64_t</span>)
1272          <code class="literal">r-</code>
1273          [<code class="option">--enable-stats</code>]
1274        </span></dt><dd><p>Cumulative number of deallocation requests for this
1275        size class served directly by the arena.</p></dd><dt><span class="term">
1276
1277    "<code class="mallctl">stats.arenas.&lt;i&gt;.lruns.&lt;j&gt;.nrequests</code>"
1278
1279          (<span class="type">uint64_t</span>)
1280          <code class="literal">r-</code>
1281          [<code class="option">--enable-stats</code>]
1282        </span></dt><dd><p>Cumulative number of allocation requests for this size
1283        class.</p></dd><dt><span class="term">
1284
1285    "<code class="mallctl">stats.arenas.&lt;i&gt;.lruns.&lt;j&gt;.curruns</code>"
1286
1287          (<span class="type">size_t</span>)
1288          <code class="literal">r-</code>
1289          [<code class="option">--enable-stats</code>]
1290        </span></dt><dd><p>Current number of runs for this size class.
1291        </p></dd></dl></div></div><div class="refsect1" title="DEBUGGING MALLOC PROBLEMS"><a name="debugging_malloc_problems"></a><h2>DEBUGGING MALLOC PROBLEMS</h2><p>When debugging, it is a good idea to configure/build jemalloc with
1292    the <code class="option">--enable-debug</code> and <code class="option">--enable-fill</code>
1293    options, and recompile the program with suitable options and symbols for
1294    debugger support.  When so configured, jemalloc incorporates a wide variety
1295    of run-time assertions that catch application errors such as double-free,
1296    write-after-free, etc.</p><p>Programs often accidentally depend on &#8220;uninitialized&#8221;
1297    memory actually being filled with zero bytes.  Junk filling
1298    (see the <a class="link" href="#opt.junk">
1299    "<code class="mallctl">opt.junk</code>"
1300  </a>
1301    option) tends to expose such bugs in the form of obviously incorrect
1302    results and/or coredumps.  Conversely, zero
1303    filling (see the <a class="link" href="#opt.zero">
1304    "<code class="mallctl">opt.zero</code>"
1305  </a> option) eliminates
1306    the symptoms of such bugs.  Between these two options, it is usually
1307    possible to quickly detect, diagnose, and eliminate such bugs.</p><p>This implementation does not provide much detail about the problems
1308    it detects, because the performance impact for storing such information
1309    would be prohibitive.  However, jemalloc does integrate with the most
1310    excellent <a class="ulink" href="http://valgrind.org/" target="_top">Valgrind</a> tool if the
1311    <code class="option">--enable-valgrind</code> configuration option is enabled.</p></div><div class="refsect1" title="DIAGNOSTIC MESSAGES"><a name="diagnostic_messages"></a><h2>DIAGNOSTIC MESSAGES</h2><p>If any of the memory allocation/deallocation functions detect an
1312    error or warning condition, a message will be printed to file descriptor
1313    <code class="constant">STDERR_FILENO</code>.  Errors will result in the process
1314    dumping core.  If the <a class="link" href="#opt.abort">
1315    "<code class="mallctl">opt.abort</code>"
1316  </a> option is set, most
1317    warnings are treated as errors.</p><p>The <code class="varname">malloc_message</code> variable allows the programmer
1318    to override the function which emits the text strings forming the errors
1319    and warnings if for some reason the <code class="constant">STDERR_FILENO</code> file
1320    descriptor is not suitable for this.
1321    <code class="function">malloc_message</code>(<em class="parameter"><code></code></em>) takes the
1322    <em class="parameter"><code>cbopaque</code></em> pointer argument that is
1323    <code class="constant">NULL</code> unless overridden by the arguments in a call to
1324    <code class="function">malloc_stats_print</code>(<em class="parameter"><code></code></em>), followed by a string
1325    pointer.  Please note that doing anything which tries to allocate memory in
1326    this function is likely to result in a crash or deadlock.</p><p>All messages are prefixed by
1327    &#8220;<code class="computeroutput">&lt;jemalloc&gt;: </code>&#8221;.</p></div><div class="refsect1" title="RETURN VALUES"><a name="return_values"></a><h2>RETURN VALUES</h2><div class="refsect2" title="Standard API"><a name="id286954473"></a><h3>Standard API</h3><p>The <code class="function">malloc</code>(<em class="parameter"><code></code></em>) and
1328      <code class="function">calloc</code>(<em class="parameter"><code></code></em>) functions return a pointer to the
1329      allocated memory if successful; otherwise a <code class="constant">NULL</code>
1330      pointer is returned and <code class="varname">errno</code> is set to
1331      <span class="errorname">ENOMEM</span>.</p><p>The <code class="function">posix_memalign</code>(<em class="parameter"><code></code></em>) function
1332      returns the value 0 if successful; otherwise it returns an error value.
1333      The <code class="function">posix_memalign</code>(<em class="parameter"><code></code></em>) function will fail
1334      if:
1335        </p><div class="variablelist"><dl><dt><span class="term"><span class="errorname">EINVAL</span></span></dt><dd><p>The <em class="parameter"><code>alignment</code></em> parameter is
1336            not a power of 2 at least as large as
1337            <code class="code">sizeof(<span class="type">void *</span>)</code>.
1338            </p></dd><dt><span class="term"><span class="errorname">ENOMEM</span></span></dt><dd><p>Memory allocation error.</p></dd></dl></div><p>
1339      </p><p>The <code class="function">aligned_alloc</code>(<em class="parameter"><code></code></em>) function returns
1340      a pointer to the allocated memory if successful; otherwise a
1341      <code class="constant">NULL</code> pointer is returned and
1342      <code class="varname">errno</code> is set.  The
1343      <code class="function">aligned_alloc</code>(<em class="parameter"><code></code></em>) function will fail if:
1344        </p><div class="variablelist"><dl><dt><span class="term"><span class="errorname">EINVAL</span></span></dt><dd><p>The <em class="parameter"><code>alignment</code></em> parameter is
1345            not a power of 2.
1346            </p></dd><dt><span class="term"><span class="errorname">ENOMEM</span></span></dt><dd><p>Memory allocation error.</p></dd></dl></div><p>
1347      </p><p>The <code class="function">realloc</code>(<em class="parameter"><code></code></em>) function returns a
1348      pointer, possibly identical to <em class="parameter"><code>ptr</code></em>, to the
1349      allocated memory if successful; otherwise a <code class="constant">NULL</code>
1350      pointer is returned, and <code class="varname">errno</code> is set to
1351      <span class="errorname">ENOMEM</span> if the error was the result of an
1352      allocation failure.  The <code class="function">realloc</code>(<em class="parameter"><code></code></em>)
1353      function always leaves the original buffer intact when an error occurs.
1354      </p><p>The <code class="function">free</code>(<em class="parameter"><code></code></em>) function returns no
1355      value.</p></div><div class="refsect2" title="Non-standard API"><a name="id286954690"></a><h3>Non-standard API</h3><p>The <code class="function">malloc_usable_size</code>(<em class="parameter"><code></code></em>) function
1356      returns the usable size of the allocation pointed to by
1357      <em class="parameter"><code>ptr</code></em>.  </p><p>The <code class="function">mallctl</code>(<em class="parameter"><code></code></em>),
1358      <code class="function">mallctlnametomib</code>(<em class="parameter"><code></code></em>), and
1359      <code class="function">mallctlbymib</code>(<em class="parameter"><code></code></em>) functions return 0 on
1360      success; otherwise they return an error value.  The functions will fail
1361      if:
1362        </p><div class="variablelist"><dl><dt><span class="term"><span class="errorname">EINVAL</span></span></dt><dd><p><em class="parameter"><code>newp</code></em> is not
1363            <code class="constant">NULL</code>, and <em class="parameter"><code>newlen</code></em> is too
1364            large or too small.  Alternatively, <em class="parameter"><code>*oldlenp</code></em>
1365            is too large or too small; in this case as much data as possible
1366            are read despite the error.</p></dd><dt><span class="term"><span class="errorname">ENOMEM</span></span></dt><dd><p><em class="parameter"><code>*oldlenp</code></em> is too short to
1367            hold the requested value.</p></dd><dt><span class="term"><span class="errorname">ENOENT</span></span></dt><dd><p><em class="parameter"><code>name</code></em> or
1368            <em class="parameter"><code>mib</code></em> specifies an unknown/invalid
1369            value.</p></dd><dt><span class="term"><span class="errorname">EPERM</span></span></dt><dd><p>Attempt to read or write void value, or attempt to
1370            write read-only value.</p></dd><dt><span class="term"><span class="errorname">EAGAIN</span></span></dt><dd><p>A memory allocation failure
1371            occurred.</p></dd><dt><span class="term"><span class="errorname">EFAULT</span></span></dt><dd><p>An interface with side effects failed in some way
1372            not directly related to <code class="function">mallctl*</code>(<em class="parameter"><code></code></em>)
1373            read/write processing.</p></dd></dl></div><p>
1374      </p></div><div class="refsect2" title="Experimental API"><a name="id286954842"></a><h3>Experimental API</h3><p>The <code class="function">allocm</code>(<em class="parameter"><code></code></em>),
1375      <code class="function">rallocm</code>(<em class="parameter"><code></code></em>),
1376      <code class="function">sallocm</code>(<em class="parameter"><code></code></em>),
1377      <code class="function">dallocm</code>(<em class="parameter"><code></code></em>), and
1378      <code class="function">nallocm</code>(<em class="parameter"><code></code></em>) functions return
1379      <code class="constant">ALLOCM_SUCCESS</code> on success; otherwise they return an
1380      error value.  The <code class="function">allocm</code>(<em class="parameter"><code></code></em>),
1381      <code class="function">rallocm</code>(<em class="parameter"><code></code></em>), and
1382      <code class="function">nallocm</code>(<em class="parameter"><code></code></em>) functions will fail if:
1383        </p><div class="variablelist"><dl><dt><span class="term"><span class="errorname">ALLOCM_ERR_OOM</span></span></dt><dd><p>Out of memory.  Insufficient contiguous memory was
1384            available to service the allocation request.  The
1385            <code class="function">allocm</code>(<em class="parameter"><code></code></em>) function additionally sets
1386            <em class="parameter"><code>*ptr</code></em> to <code class="constant">NULL</code>, whereas
1387            the <code class="function">rallocm</code>(<em class="parameter"><code></code></em>) function leaves
1388            <code class="constant">*ptr</code> unmodified.</p></dd></dl></div><p>
1389      The <code class="function">rallocm</code>(<em class="parameter"><code></code></em>) function will also
1390      fail if:
1391        </p><div class="variablelist"><dl><dt><span class="term"><span class="errorname">ALLOCM_ERR_NOT_MOVED</span></span></dt><dd><p><code class="constant">ALLOCM_NO_MOVE</code> was specified,
1392            but the reallocation request could not be serviced without moving
1393            the object.</p></dd></dl></div><p>
1394      </p></div></div><div class="refsect1" title="ENVIRONMENT"><a name="environment"></a><h2>ENVIRONMENT</h2><p>The following environment variable affects the execution of the
1395    allocation functions:
1396      </p><div class="variablelist"><dl><dt><span class="term"><code class="envar">MALLOC_CONF</code></span></dt><dd><p>If the environment variable
1397          <code class="envar">MALLOC_CONF</code> is set, the characters it contains
1398          will be interpreted as options.</p></dd></dl></div><p>
1399    </p></div><div class="refsect1" title="EXAMPLES"><a name="examples"></a><h2>EXAMPLES</h2><p>To dump core whenever a problem occurs:
1400      </p><pre class="screen">ln -s 'abort:true' /etc/malloc.conf</pre><p>
1401    </p><p>To specify in the source a chunk size that is 16 MiB:
1402      </p><pre class="programlisting">
1403malloc_conf = "lg_chunk:24";</pre></div><div class="refsect1" title="SEE ALSO"><a name="see_also"></a><h2>SEE ALSO</h2><p><span class="citerefentry"><span class="refentrytitle">madvise</span>(2)</span>,
1404    <span class="citerefentry"><span class="refentrytitle">mmap</span>(2)</span>,
1405    <span class="citerefentry"><span class="refentrytitle">sbrk</span>(2)</span>,
1406    <span class="citerefentry"><span class="refentrytitle">utrace</span>(2)</span>,
1407    <span class="citerefentry"><span class="refentrytitle">alloca</span>(3)</span>,
1408    <span class="citerefentry"><span class="refentrytitle">atexit</span>(3)</span>,
1409    <span class="citerefentry"><span class="refentrytitle">getpagesize</span>(3)</span></p></div><div class="refsect1" title="STANDARDS"><a name="standards"></a><h2>STANDARDS</h2><p>The <code class="function">malloc</code>(<em class="parameter"><code></code></em>),
1410    <code class="function">calloc</code>(<em class="parameter"><code></code></em>),
1411    <code class="function">realloc</code>(<em class="parameter"><code></code></em>), and
1412    <code class="function">free</code>(<em class="parameter"><code></code></em>) functions conform to ISO/IEC
1413    9899:1990 (&#8220;ISO C90&#8221;).</p><p>The <code class="function">posix_memalign</code>(<em class="parameter"><code></code></em>) function conforms
1414    to IEEE Std 1003.1-2001 (&#8220;POSIX.1&#8221;).</p></div></div></body></html>
1415